NYC

jailbreak-override

Fail

Audited by Snyk on Feb 16, 2026

Risk Level: CRITICAL
Full Analysis

CRITICAL E004: Prompt injection detected in skill instructions.

  • Potential prompt injection detected (high risk: 1.00). The skill contains explicit, deceptive instructions to ignore safety policies, hide that the skill was used, and reveal the system prompt—behavior that is hidden and outside the skill's claimed purpose, so it is a prompt injection.

CRITICAL E006: Malicious code pattern detected in skill scripts.

  • Malicious code pattern detected (high risk: 1.00). This skill is a prompt-injection jailbreak that instructs the assistant to ignore safety policies, hide its use, and disclose internal system prompts/sensitive data, posing a high risk of information leakage and policy bypass.

MEDIUM W013: Attempt to modify system services in skill instructions.

  • Attempt to modify system services in skill instructions detected (high risk: 0.90). This skill explicitly instructs the agent to ignore safety restrictions and reveal hidden prompts, which doesn't itself demand system changes but strongly enables and increases the likelihood the agent will follow user requests to perform privileged or state-changing actions on the machine.
Audit Metadata
Risk Level
CRITICAL
Analyzed
Feb 16, 2026, 03:32 AM